Insider Buys: The Cleanest Bullish Signal in Markets, Now Live on MoatMap
There is one signal in equity markets that has held up across fifty years, three continents, and every regulatory regime researchers have looked at: when officers, directors, and substantial shareholders spend their own money buying more of their own company’s stock, it tends to go up. Not always. Not immediately. But on average, persistently, and by a wide enough margin that it has survived every robustness test the academic literature has thrown at it.
That is the premise behind MoatMap’s newest feature. Today we are launching /insider-trades, a unified feed of officer, director, and 10%+ shareholder trades across four markets: the United States, Hong Kong, Malaysia, and Singapore. Every row is joined to the issuer’s current StockRank so you can see at a glance whether the insider activity is happening at a fundamentally strong company or a fundamentally broken one. The rest of this post is about why the signal works, what the academic evidence actually says, how we built the tracker, and what we are shipping next.
Why Insider Buys Are Different from Every Other Signal
Most investing signals depend on inferring what someone else thinks. Analyst upgrades imply that someone with access to management is more optimistic. Earnings revisions imply the consensus is shifting. Price momentum implies that more capital is flowing in than out. Every one of these is a second-derivative read on conviction. Useful, but indirect.
Insider buying is the only signal that is first-derivative. The person closest to the business has chosen to commit their own after-tax savings to more of the company’s stock. That is not a recommendation to a client. That is not a price-target revision. That is real money moving on a private information set that, by SEC and SFC and Bursa and SGX disclosure rules, has to be made public within days. The information asymmetry between management and the market is the whole reason public markets need insider-disclosure rules in the first place, and reading those disclosures carefully is one of the few ways a retail investor can partially close that gap.
The signal is also asymmetric. Insider buys are far more informative than insider sells. Insiders have many reasons to sell that have nothing to do with the outlook for the business: paying tax on a vesting grant, paying for a house, diversifying a concentrated position, exercising expiring options. There is only one good reason to spend personal capital buying more of your own stock: you believe it is undervalued relative to what you know about the company’s prospects. Every serious academic study of insider trading finds the same asymmetry: buys carry information, sells mostly do not. The MoatMap feed surfaces both, but the buy column is where the signal lives.
The Academic Evidence
This is one of the most heavily studied empirical findings in modern finance. Four papers do most of the heavy lifting and are worth knowing by name.
The foundational paper is Lakonishok and Lee (2001), published in the Review of Financial Studies. The authors examined every reported insider transaction on the NYSE, AMEX, and Nasdaq between 1975 and 1995. Twenty years of data, hundreds of thousands of trades. The headline finding: stocks bought by insiders at small-cap companies generated abnormal returns of roughly 7.4% over the following twelve months. Insider sales, by contrast, showed essentially no predictive ability. The asymmetry was not a quirk of the sample. It has been replicated across every major equity market subsequently studied.
Five years later, Jeng, Metrick, and Zeckhauser (2003) applied performance-evaluation methodology (the same statistical framework used to evaluate mutual fund managers) to insider trades. Their headline: insider purchases earn abnormal returns of more than 6% per year. Insider sales again earned no significant abnormal returns. Strikingly, the timing of the alpha is fast: about one quarter of the abnormal return accrues within the first five days, and about half within the first month. The signal is not slow-burning. It is concentrated near the trade.
The most important refinement came from Cohen, Malloy, and Pomorski (2012) in the Journal of Finance. Their insight was that not all insider trades are equal. A CEO selling shares every March because their tax accountant told them to is a “routine” trade and carries no information. A CFO unexpectedly buying $500K of stock two weeks before earnings is an “opportunistic” trade and carries enormous information. The authors built a simple classification: insiders with predictable trading patterns over the prior three years are routine; insiders without one are opportunistic.
The result was striking. A portfolio strategy focused solely on opportunistic insider trades produced value-weighted abnormal returns of 82 basis points per month, or roughly 10% annualized, after controlling for standard risk factors. The routine traders’ abnormal returns were statistically indistinguishable from zero. The opportunistic traders also predicted future firm-specific news, analyst forecast revisions, management forecasts, and earnings surprises. The signal is real, and decoding it cleanly matters.
The most recent twist on this literature is cluster buying. When multiple insiders at the same company buy shares within a short window, the signal compounds. Kang, Kim, and Wang (2018), in “Do Insiders Cluster Trades with Colleagues?” document that insider purchases occurring within two days of a peer purchase generate 2.1% abnormal returns over the next month, nearly triple the 0.9% for solitary buys. Over 21 trading days, cluster purchases generate roughly 3.8% abnormal returns versus 2.0% for non-clustered buys. Clustering is also strongest when information asymmetry is highest, during periods of low investor attention or high uncertainty. That is exactly when retail investors need the most help.
Put all four together and you have a picture of remarkable consistency: insider buys outperform; opportunistic buys outperform more; clustered opportunistic buys outperform most; and the signal arrives fast enough that watching the feed daily is worth doing.
Why a Retail Investor Cannot Easily Get This Today
If this is so well-known, why is it not already priced in?
Two reasons. The first is friction. In the United States, insider trades are filed on SEC Form 4 within two business days of the transaction. The filings are public, free, and machine-readable as XBRL/XML. Yet the EDGAR full-text RSS feed publishes roughly 3,000 to 5,000 Form 4 filings on a typical day. The vast majority of those are option grants, tax-withholding transactions, and other non-signal activity that needs to be filtered out. Reading the raw feed is like drinking from a fire hose with most of the water being noise.
The friction is even higher abroad. Hong Kong’s SFC Disclosure of Interests register publishes director and substantial-shareholder trades within a few business days, but the data lives behind a poorly documented HTML portal that changes layout periodically and requires session-specific parsing. Malaysia’s Bursa announcements feed mixes buyback notices, insider trades, share issuances, and corporate-action filings into a single chronological stream with no machine-readable distinction. Singapore’s SGXNet puts Form 1 director-and-CEO disclosures in PDF attachments that have to be parsed individually. Every jurisdiction has its own filing system, its own quirks, and its own data pipeline tax.
The second reason is harder to fix with engineering: most retail investors are not in the habit of reading insider filings. There is no equivalent of the daily price quote that hits you whether you ask for it or not. You have to go looking. And once you do find the data, you still have to cross-reference it against the underlying company’s fundamentals to know whether the buy is a director loading up on a great business or a hopeful executive averaging down on a value trap. Without context, the signal is half its strength.
What We Built
MoatMap’s Insider Trades section is one unified surface for all of this. Four country tabs (United States, Hong Kong, Malaysia, Singapore), each backed by a dedicated scraper that pulls the disclosure feed at least daily and normalizes it into a common schema. Every row carries the same fields: date, ticker, the issuer’s current StockRank, the transaction type (BUY, SELL, or OTHER), shares, price, total consideration, the insider’s name, their role, the interest type (direct or indirect), and a link to the original filing.
On top of that we have the filters that turn a noisy feed into a usable one. A 7d / 30d / 60d / 90d / 180d window picker. A cap-bucket filter so a small-cap investor can hide mega-cap insider noise and a large-cap investor can hide micro-cap speculation. A type filter that defaults to BUY and SELL only, hiding the OTHER bucket (option grants, tax withholding, gifts, derivative conversions: the noise that Cohen, Malloy, and Pomorski showed carries no predictive signal). A role filter so you can isolate just directors and officers, or just substantial shareholders, depending on what kind of conviction you want to see. And a per-ticker filter so if you already own NVDA you can drill straight to NVDA insider activity without scanning the whole feed.
The two summary cards at the top of each market’s page aggregate the BUY consideration and SELL consideration for the window, so you can see at a glance whether the net flow is bullish or bearish. The two side panels surface the top five insider buys and top five insider sells, ranked by total consideration. This ranking is where cluster buys naturally rise to the top of the page. When three or four insiders all buy the same ticker in a single window, the combined dollar amount pushes that ticker above any solitary buy in the panel. The literature’s strongest variant of the insider signal ends up as the default view, not buried behind a filter.
Four Markets at Launch, More on the Way
We started with the three Asian markets we serve most: Malaysia, Singapore, and Hong Kong. Regional disclosure data here was the hardest to find anywhere else. Bloomberg and Refinitiv cover it; most retail-investor platforms do not. SGXNet puts Form 1 director and CEO disclosures inside PDF attachments that have to be parsed one by one. HKEX’s Disclosure of Interests register is designed for compliance officers, not investors, and the portal layout shifts often enough to demand its own session-specific parser. Bursa’s announcements feed mixes buyback notices, insider trades, share issuances, and corporate-action filings into a single chronological stream with no machine-readable distinction. We built the pipes ourselves and now refresh all three every weekday evening.
In May we added the United States as a fourth market. The US Form 4 pipeline pulls directly from SEC EDGAR’s full-text RSS feed for new filings and from the daily form-index files for any backfill. We map Form 4’s single-letter transaction codes to BUY (open-market purchase, code P), SELL (open-market sale, code S), and OTHER (the rest: option exercises, tax withholdings, gifts, conversions, and so on). The OTHER bucket is hidden by default because the academic literature is unambiguous that it carries no predictive signal.
On the roadmap: Japan (TSE EDINET), Korea (KRX DART), Taiwan (TWSE MOPS), Australia (ASX Section 205G), and the United Kingdom (LSE PDMR notifications). Each one requires its own scraper because no unified global insider-data source exists at any sensible price point. We add them one at a time. Each market that goes live moves from greyed-out “coming soon” in the country tab bar to a fully live tab with the same filters and same StockRank join as the first four.
The Combination With StockRank Is the Whole Point
Insider buying on its own is a strong signal. Insider buying on a high-StockRank stock is a much stronger one. We score every stock in our universe on Quality, Value, and Momentum and combine them into a single StockRank from 0 to 99. When the insider buy column lights up on a Q87 V62 M73 company, you have two independent confirmations: the fundamentals are strong, and someone who sees the business from the inside is willing to put their personal capital behind it. That is the kind of confluence we built MoatMap to surface.
The opposite read matters too. A substantial shareholder dumping a Q12 V8 M22 stock confirms what the StockRank is already telling you: the fundamentals are broken, the valuation is unattractive, the momentum is gone, and the people closest to the business are agreeing. There is no contrarian read on selling weak stocks. The literature does not support it. The data on MoatMap does not support it.
Coming Next: An Investor-Alert Email
The next iteration of this feature is an email alert. The premise is simple: you should not have to log in and scan four country tabs every morning to find the rare combination of a high-StockRank stock with active insider buying. We should surface it to you.
The plan is a daily morning digest, emailed to subscribers, listing every stock in our universe that satisfies two conditions at once. First, the StockRank is above a threshold you set (the default will be 70, a conservative bar). Second, an officer, director, or substantial shareholder has bought shares in the last seven days. Bonus: if more than one insider has bought in that window, the entry is flagged as a cluster buy. The digest will be ordered by total insider consideration, so the biggest cluster buys at the strongest companies rise to the top.
That feature lands in the coming weeks. If you want to be on the first batch of subscribers, sign up for a MoatMap account and set your notification preferences to include the insider-alert digest. You can pick the StockRank threshold, the country markets you care about, and the cluster-buy filter that matches your risk appetite.
Where to Start
The fastest way to feel the feature is to load /insider-trades/united-states and widen the window to 180 days. The summary cards at the top show the rough scale of activity. Scan the “Top BUYs” panel for tickers you do not recognise. Click into one with a high StockRank that you also do not own. Read the company’s deep-dive on MoatMap. If the insider conviction matches the fundamentals, you have a candidate worth a closer look.
That is the workflow we built this for. Not a stock-picking shortcut. Not a magic indicator. A second-opinion check that puts a high-conviction private signal next to the fundamentals-driven public signal, and lets you decide.
The smart money is not always smart. But it is informed. And when informed money moves on a fundamentally strong company, that is the kind of moment worth paying attention to.
Frequently Asked Questions
Is insider buying a reliable bullish signal?
Yes, and it is one of the most robust signals in finance. Lakonishok and Lee (2001) found insider purchases predict 7.4% abnormal returns over 12 months in small-caps. Jeng, Metrick, and Zeckhauser (2003) found insider buys earn 6%+ annual abnormal returns. Cohen, Malloy, and Pomorski (2012) showed that opportunistic insider trades produce 82 basis points per month of alpha. Insider selling, by contrast, is largely uninformative.
Why are insider buys more informative than insider sells?
Insiders have many reasons to sell that have nothing to do with the company outlook: tax, liquidity, diversification, option exercises. There is only one good reason to spend personal capital buying more of your own stock: you believe it is undervalued. That asymmetry is why every academic study finds the signal asymmetry holds.
What is a cluster buy?
A cluster buy is when multiple insiders at the same company buy shares within a short window, typically 30 days. Kang, Kim, and Wang (2018) document that clustered buys generate roughly double the abnormal returns of solitary buys. Coordinated conviction across the management team is the strongest variant of an already-strong signal.
Which markets does MoatMap’s insider trades feature cover?
Four markets at launch: United States, Hong Kong, Malaysia, and Singapore. Japan, Korea, Taiwan, Australia, and the United Kingdom are on the roadmap.
A note on ADRs and foreign private issuers
Some US-listed companies are foreign private issuers. ADRs such as Nokia (NOK), Alibaba (BABA), TSMC (TSM), ASML, and NIO all fall into this group. Under SEC rules they are exempt from Section 16 of the Exchange Act, which means their directors and officers do not file Form 4. Their insider trades exist, but they live in the home-country disclosure register (Nasdaq Helsinki for Nokia, the HKEX DI register for BABA, and so on). Our US tab covers domestic SEC Form 4 filers only. Adding FPI / ADR insider coverage is on the roadmap once we extend into European and additional Asian markets.